tyktechnologies/tyk-install
Deploy, verify, and production-harden Tyk with an AI coding agent. Plans a deployment topology across Docker, Kubernetes/Helm, OpenShift, and bare metal for the open source Gateway (no licence), Self-Managed, hybrid, MDCB and AI Studio, runs it, proves it works with executable evidence, and applies the production and platform guidance this repository already carries as annotated config.
Choose and prepare a Tyk deployment before anything is created. Use when someone wants to install, deploy, evaluate, or stand up Tyk and the topology is not already decided — picking between OSS, Self-Managed, hybrid, MDCB, or AI Studio, between Docker, Kubernetes/Helm, OpenShift, or bare metal, and between a local trial and a production install. Checks prerequisites, resolves licences and versions, generates configuration overlays, and writes a deployment plan. Produces a plan only; it starts nothing.
Start a Tyk deployment, bootstrap it, and prove it works. Use when running or restarting a Tyk stack with Docker Compose, Helm, or bare-metal packages, when bootstrapping an organisation, Dashboard user, or first API, and when a Tyk deployment is up but not behaving — gateway not starting, Dashboard licence errors, Redis or Postgres connection failures, MDCB RPC handshake problems, or a request not reaching its upstream. Collects executable evidence rather than reporting success from container status alone.
Audit or harden a Tyk configuration against production guidance. Use when asked whether a Tyk deployment is production ready, to review a values.yaml or gateway env file before going live, to apply performance or scale tuning, to check Redis and Postgres placement, TLS, secrets, rate limiting, or analytics storage, or to prepare an OpenShift overlay. Reads the production, performance, security, and OpenShift guidance this repository already carries as annotated config, so recommendations cannot drift from the configuration they describe.